Kalpa Labs
Scaling Generalist Speech models
Y CombinatorFall 2025ActiveB2B
What they do
Problem
Current speech AI systems are fragmented and lack generalist capabilities like strong instruction-following and in-context learning across diverse speech tasks.
Solution
Kalpa Labs builds scalable generalist speech foundation models that unify speech-to-text, text-to-speech, voice cloning, and instruction-following/in-context learning in a single model.
Who it's for
Developers and AI product teams building voice-enabled applications and speech-driven agents.
Founders
Prashant Shishodia
Founder
Pushing frontier of speech models @ KalpaLabs. Previously led full stack ML @ Google scaling to billions of queries / month.
Gautam Jha
Founder
Pushing frontier of speech models @ KalpaLabs. Previously built nanoseconds latency software at HFTs.
Open roles 1
Founding ML Research Engineer - Training Infrastructure
Company timeline
- Founded2025
- YC batchFall 2025
- StatusActive
- Team size2 people
- LocationSan Francisco, CA, USA
Similar startups
Lyrebird.ai
Lyrebird offers a new generation of speech synthesis technologies for…
Speechly
Natural Language Understanding API for Speech that runs on device
Miso Labs
The most emotive foundation models for voice
Vapi
Voice AI for developers.
Deepgram
Building foundational AI for speech transcription and understanding.
Pretzel AI
AI tooling